Limited value of current shoulder arthroplasty registries in evidence-based shoulder surgery: a review of 7 national registries.

National shoulder arthroplasty registries are currently used to assess incidence, indication, type of prosthesis and revision, but they seem to lack sufficient information to lead to evidence based decision-making in shoulder surgery. There appears to be a large difference in registered parameters and outcome measurement per country. First we investigated whether existing registries have sufficient common datasets to enable pooling of data. Second, we determined whether known risk factors for prosthetic failure are being recorded. Through a non-systematic literature review studies on registries were analyzed for included parameters. Seven national registries were scrutinized for the data collected and these were classified according to categories of risk factors for failure: patient-, implant and surgeon related, and other parameters. This shows a large heterogeneity of registered parameters between countries. The majority of parameters shown to be relevant to outcome and failure of shoulder prostheses are not included in the studied registries. International agreement on parameters and outcome measurement for registries is paramount to enable pooling and comparison of data. If we intend to use the registries to provide us with evidence to improve prosthetic shoulder surgery, we need adjustment of the different parameters to be included.

Clavicular osteotomy in complicated revision of total shoulder arthroplasty: indications, surgical technique, and outcomes.

PURPOSE: Adequate exposure in revision of total shoulder arthroplasty (TSA) is important for optimal prosthesis placement and functional results. A clavicular osteotomy in difficult cases of revision TSA is a useful surgical technique that increases the superior exposure area, provides safer dissection, minimizes damage to the anterior deltoid muscle, improves glenoid access, and allows for superior dislocation of the humeral component. There is a paucity of literature analyzing the clavicular osteotomy during challenging cases of revision TSA. The aims of this study were to describe the application, surgical technique, and outcomes of revision TSA with a clavicular osteotomy. METHODS: This was a retrospective study of consecutive patients who underwent revision TSA with a clavicle osteotomy at a single institution (2004-2016). A curved longitudinal clavicular osteotomy is created parallel to the origin of the anterior deltoid muscle. This allows for lateral reflection of the osteotomy and anterior deltoid muscle to significantly increase superior exposure and reduce damage to remaining deltoid muscle fibres. Osteotomy closure is simple with four or five Nice knot osteosutures. The Constant-Murley score and osteotomy healing were assessed at every follow-up. All complications were reviewed. RESULTS: Forty patients who had a mean age of 63.8 years (range 37-87) at time of surgery and mean follow-up duration of 34 months (range 12-88) were analyzed. Pre-operative Constant-Murley scores improved significantly from 32 ± 19.0 to 58 ± 15.0 (p < 0.001) at one year and 65 ± 13.1 (p < 0.001) at two years. Primary osteotomy healing and callus formation were evident in 95% of cases by three months. Five patients developed post-operative complications (13%) related to the clavicular osteotomy: three mid-diaphyseal clavicular fractures sustained after trauma (8%), one clavicular stress fracture (3%), and case of one loosening (3%). Three patients (8%) required surgical revision of the osteotomy (two internal fixation and one revision osteosuturing). No neurovascular injuries or scapular fractures were encountered. CONCLUSION: A curved longitudinal clavicular osteotomy is beneficial in difficult revision TSA and is another tool in the arsenal of experienced shoulder surgeons who manage these challenging cases. This surgical technique increases glenoid exposure, facilitates superior dislocation of the humeral component, minimizes anterior deltoid damage, and reduces the risk of neurovascular injuries. All clavicular complications occurred within four months prior to osteotomy union, with many sustained due to trauma. However, patients who developed a complication had comparable shoulder function as those without.

Morphologic variations of the scapula in 3-dimensions: a statistical shape model approach.

BACKGROUND: Morphologic variations of the scapula and acromion have been found to be associated with shoulder pathology. This study used statistical shape modelling to quantify these variations in healthy shoulders. MATERIALS AND METHODS: A statistical shape model of the scapula was created using 3-dimensional computed tomography reconstructions of 108 survey-confirmed nonpathologic shoulders of 54 patients. The mean shape and the 95% confidence interval were calculated and analyzed in the first 5 shape modes. RESULTS: The first 5 shape modes consisted of consecutively sized (72% of total variation), rotation of the coracoacromial complex (5%), acromial shape and slope (4%), shape of the scapular spine (2%), and acromial overhang (2%). DISCUSSION AND CONCLUSION: In healthy shoulders, a certain variation in rotation of the coracoacromial complex and in acromial shape and slope was observed. These new parameters might be correlated with shoulder pathology such as glenohumeral osteoarthritis or rotator cuff tears.

Fracture of the proximal humerus with disruption of the tendon of the pectoralis major.

Combined pectoralis major disruption and proximal humeral fractures are uncommon. A simple radiologic diagnostic tool which consists of the measurement of the displacement from the humeral shaft to the lateral side of the humeral head (lateral to the outer proximal cortex) can help to diagnose this combined lesion.

Can an extracorporeal glenoid aiming device be used to optimize the position of the glenoid component in total shoulder arthroplasty?

PURPOSE: Successful total shoulder arthroplasty (TSA) requires a correct position of the glenoid component. This study compares the accuracy of the positioning with a new developed glenoid aiming device and virtual three-dimensional computed tomography (3D-CT) scan positioning. MATERIALS AND METHODS: On 39 scapulas from cadavers, a K-wire (KDev) was positioned using the glenoid aiming device. It consists of glenoid components connected to the aiming device, which cover 150° of the inferior glenoid circle, has a fixed version and inclination and is available with several different radii. The aiming device is stabilized at the most medial scapular point. The K-wire is drilled from the center of the glenoid component to this most medial point. All scapulas were also scanned with CT and 3D reconstructed. A virtual K-wire (Kct) was positioned in the center of the glenoid and in the scapular plane. Several parameters were compared. Radius of the chosen glenoid component (rDev) and the virtual radius of the glenoid circle (rCT), spinal scapular length with the device (SSLdev) and virtual (SSLct), version and inclination between KDev and Kct, difference between entry point and exit point ("Matsen"-point). RESULTS: Mean rDev: 14 mm ± 1.7 mm and mean rCT: 13.5 mm ± 1.6 mm. There was no significant difference between SSLdev (110.6 mm ± 7.5 mm) and SSLct (108 mm ± 7.5 mm). The version of KDev and Kct was -2.53° and -2.17° and the inclination 111.29° and 111.66°, respectively. The distance between the "Matsen-point" device and CT was 1.8 mm. CONCLUSION: This glenoid aiming device can position the K-wire on the glenoid with great accuracy and can, therefore, be helpful to position the glenoid component in TSA. The level of evidence: II.

Consequences of reaming with flat and convex reamers for bone volume and surface area of the glenoid; a basic science study.

BACKGROUND: The effect of reaming on bone volume and surface area of the glenoid is not precisely known. We hypothesize that (1) convex reamers create a larger surface area than flat reamers, (2) flat reamers cause less bone loss than convex reamers, and (3) the amount of bone loss increases with the amount of version correction. METHODS: Reaming procedures with different types of reamers are performed on similar-sized uniconcave and biconcave glenoids created from Sawbones foam blocks. The loss of bone volume, the size of the remaining surface area, and the reaming depth are measured and evaluated. RESULTS: Reaming with convex reamers results in a significantly larger surface area than with flat reamers for both uniconcave and biconcave glenoids (p = 0.013 and p = 0.001). Convex reamers cause more bone loss than flat reamers, but the difference is only significant for uniconcave glenoids (p = 0.007). CONCLUSIONS: In biconcave glenoids, convex reamers remove a similar amount of bone as flat reamers, but offer a larger surface area while maximizing the correction of the retroversion. In pathological uniconcave glenoids, convex reamers are preferred because of the conforming shape.

Rocking-horse phenomenon of the glenoid component: the importance of inclination.

BACKGROUND: Abnormal glenoid version positioning has been recognized as a cause of glenoid component failure caused by the rocking horse phenomenon. In contrast, the importance of the glenoid inclination has not been investigated. MATERIALS AND METHODS: The computed tomography scans of 152 healthy shoulders were evaluated. A virtual glenoid component was positioned in 2 different planes: the maximum circular plane (MCP) and the inferior circle plane (ICP). The MCP was defined by the best fitting circle of the most superior point of the glenoid and 2 points at the lower glenoid rim. The ICP was defined by the best fitting circle on the rim of the inferior quadrants. The inclination of both planes was measured as the intersection with the scapular plane. We defined the force vector of the rotator force couple and calculated the magnitude of the shear force vector on a virtual glenoid component in both planes during glenohumeral abduction. RESULTS: The inclination of the component positioned in the MCP averaged 95° (range, 84°-108°) and for the ICP averaged 111° (range, 94°-126°). A significant reduction in shear forces was calculated for the glenoid component in the ICP vs the MCP: 98% reduction in 60° of abduction to 49% reduction in 90° of abduction. CONCLUSION: Shear forces are significantly higher when the glenoid component is positioned in the MCP compared with the ICP, and this is more pronounced in early abduction. Positioning the glenoid component in the inferior circle might reduce the risk of a rocking horse phenomenon.

A glenoid reaming study: how accurate are current reaming techniques?

BACKGROUND: Correct reaming of a degenerative glenoid can be a difficult procedure. We investigated how the quality of the reamed surface is influenced by different reamers, by the surgeon's experience, and by glenoid erosion patterns. MATERIAL AND METHODS: Three shoulder surgeons performed reaming procedures with different types of reamers (flat, convex, K-wire guided, and nipple guided) on a series of similarly sized uniconcave and biconcave glenoids. The reproducibility of reaming and the effect of different reamers on different-shaped glenoids were measured and evaluated. RESULTS: The center and direction of reaming were constant for all surgeons in the case of type A glenoids. For type B2 glenoids, the center and direction of reaming differed significantly between surgeons. The congruity of the reamed surface was better after flat reaming than after convex reaming. Whether the reamers were guided by a central K-wire or by a nipple had no significant effect on the reamed surface. The experience of the surgeon had no effect on the congruity of reaming. CONCLUSIONS: Reaming of a uniconcave glenoid is reproducible, but reaming of a biconcave glenoid seems much more difficult. Erosion and deformity of the glenoid influence the accuracy of reaming the most. Surgical experience plays a less important role. We conclude that there is a need for guidance in reaming of biconcave glenoids.

Three-dimensional computed tomography scan evaluation of the pattern of erosion in type B glenoids.

BACKGROUND: Type B glenoids according to Walch are subclassified into a group with no posterior erosion (B1) and a group with important posterior erosion (B2) based on 2-dimensional axial computed tomography (CT) scan images. Three-dimensional (3D) CT scan reconstruction seems to improve the accuracy of the measurement of erosion because of its independence from positional errors. The aim is to quantify the direction and amount of posterior erosion of type B glenoids using a reproducible 3D measuring technique. METHODS: We performed 3D reconstruction of 72 type B glenoids (24 type B1 glenoids and 48 type B2 glenoids) using Mimics (Materialise, Haasrode, Belgium). The native glenoid plane and intermediate glenoid plane were determined by use of 3-Matic (Materialise). The normal glenoid version, eroded retroversion, and difference in retroversion were measured. Next, the maximum erosion and its orientation were quantified. RESULTS: There was always minimum erosion of 1.7 mm, and the mean erosion was 4.2 mm. There was a significant difference between the mean erosion in type B1 glenoids (3.5 mm) and type B2 glenoids (4.5 mm) (P = .019). The mean orientation of the erosion was mostly to the posteroinferior side (119°; SD, 26.8). There was a significant difference between the mean orientation in type B1 glenoids (132°; SD, 25.2) and type B2 glenoids (113°; SD, 25.5) (P = .004). CONCLUSIONS: With the use of this newly developed reproducible measuring technique, the maximum erosion in type B glenoids can be adequately quantified. All glenoids showed signs of important erosion. Because the orientation of the maximum erosion in type B1 glenoids is situated more inferiorly, the 2-dimensional CT scan technique can be insufficient to evaluate this erosion.

Determination of a reference system for the three-dimensional study of the glenohumeral relationship.

OBJECTIVE: Knowledge of the normal and pathological three-dimensional glenohumeral relationship is imperative when planning and performing a total shoulder arthroplasty. There is, however, no consensus on which references should be used when studying this relationship. The purpose of the present study was to define the most suitable glenoid plane with normally distributed parameters, narrowest variability, and best reproducibility. MATERIALS AND METHODS: Three-dimensional reconstruction CT scans were performed on 152 healthy shoulders. Four glenoid planes, each determined by three surgically accessible bony reference points, were determined. Two planes were triangular, with the same base defined by the most anterior and posterior point of the glenoid. The most inferior and the most superior point of the glenoid, respectively, define the top of Saller's inferior plane and the Saller's superior plane. The two other planes are formed by best-fitting circles. The circular max plane is defined by the superior tubercle, and two points at the distal third of the glenoid. The circular inferior plane is defined by three points at the rim of the inferior quadrants of the glenoid. RESULTS: The parameters of all four planes behave normally. The humeral center of rotation is identically positioned for both the circular max and circular inferior plane (X = 91.71°/X = 91.66° p = 0.907 and Y = 90.83°/Y = 91.7° p = 0.054, respectively) and different for the Saller's inferior and Saller's superior plane (p ≤ 0.001). The circular inferior plane has the lowest variability to the coronal scapular plane (p < 0.001). CONCLUSIONS: This study provides arguments to use the circular inferior glenoid plane as preferred reference plane of the glenoid.

Does adding heavy load eccentric training to rehabilitation of patients with unilateral subacromial impingement result in better outcome? A randomized, clinical trial.

PURPOSE: To investigate superior value of adding heavy load eccentric training to conservative treatment in patients with subacromial impingement. METHODS: Sixty-one patients with subacromial impingement were included and randomly allocated to the traditional rotator cuff training (TT) group (n = 30, mean age = 39.4 ± 13.1 years) or traditional rotator training combined with heavy load eccentric training (TT + ET) group (n = 31, mean age = 40.2 ± 12.9 years). Isometric strength was measured to abduction at 0°, 45° and 90° of scapular abduction and to internal and external rotation. The SPADI questionnaire was used to measure shoulder pain and function. Patients rated subjective perception of improvement. Outcome was assessed at baseline, at 6 and 12 weeks after start of the intervention. Both groups received 9 physiotherapy treatments over 12 weeks. At home, the TT group performed traditional rotator cuff strengthening exercises 1x/day. The TT + ET group performed the same exercises 1x/day and a heavy load eccentric exercise 2x/day. RESULTS: After treatment, isometric strength had significantly increased in all directions, and SPADI score had significantly decreased. The TT + ET group showed a 15% higher gain in abduction strength at 90° of scapular abduction. Chi-square tests showed patients' self-rated perception of improvement was similar in both groups. CONCLUSION: Adding heavy load eccentric training resulted in a higher gain in isometric strength at 90° of scapular abduction, but was not superior for decreasing pain and improving shoulder function. This study showed that the combination of a limited amount of physiotherapy sessions combined with a daily home exercise programme is highly effective in patients with impingement. LEVEL OF EVIDENCE: II.

Operative guidelines for the reconstruction of the native glenoid plane: an anatomic three-dimensional computed tomography-scan reconstruction study.

BACKGROUND: Reconstruction of the native plane in biconcave eroded glenoids is difficult. Nevertheless, accurate reconstruction of this plane is imperative for successful total shoulder arthroplasty. This study aims to determine guidelines that can increase the accuracy of glenoid component positioning. METHODS: Three different circular planes were determined on 3-dimensional computed tomography (CT) scans of 152 healthy shoulders. First, the circular max (CM) plane is formed with the superior tubercle and 2 points, 1 anterior and 1 posterior, at the rim of the inferior third of the glenoid. Second, the circular inferior (CI) plane is formed by 3 points at the inferior 2 quadrants of the glenoid rim. Third, the circular minima (Cm) plane is formed with 3 points situated at the noneroded sector of the anterior glenoid. The angulation of the spinal scapular axis (SSA), the line between the most medial point of the scapular spine and the center of the three different glenoid planes, and the correlation coefficient between the radius of the circle and the length of SSA are calculated. RESULTS: Angle SSA in the x-axis were 94°, 93°, 93° and in the y-axis were 95°, 111°, and 111° for CM, CI, and Cm, respectively. Correlation coefficient between the radius of the circle and the length of SSA: r = 0.69 for CM, r = 0.75 for CI, and r = 0.75 for Cm. CONCLUSION: Three points situated at the native anterior glenoid can reconstruct, within 2° accuracy (95% confidence interval, 1.8°-2.3°), the CI plane. A relationship exists between the radii of the 3 glenoid circles and the width of the scapula (SSA length).

The impact of rotator cuff tendinopathy on proprioception, measuring force sensation.

BACKGROUND: The impact of rotator cuff tendinopathy and related impingement on proprioception is not well understood. Numerous quantitative and qualitative changes in shoulder muscles have been shown in patients with rotator cuff tendinopathy. These findings suggest that control of force might be affected. This investigation wants to evaluate force sensation, a submodality of proprioception, in patients with rotator cuff tendinopathy. METHODS: Thirty-six patients with rotator cuff tendinopathy and 30 matched healthy subjects performed force reproduction tests to isometric external and internal rotation to investigate how accurately they could reproduce a fixed target (50% MVC). Relative error, constant error, and force steadiness were calculated to evaluate respectively magnitude of error made during the test, direction of this error (overshoot or undershoot), and fluctuations of produced forces. RESULTS: Patients significantly overshoot the target (mean, 6.04% of target) while healthy subjects underestimate the target (mean, -5.76% of target). Relative error and force steadiness are similar in patients with rotator cuff tendinopathy and healthy subjects. Force reproduction tests, as executed in this study, were found to be highly reliable (ICC 0.849 and 0.909). Errors were significantly larger during external rotation tests, compared to internal rotation. CONCLUSION: Patients overestimate the target during force reproduction tests. This should be taken into account in the rehabilitation of patients with rotator cuff tendinopathy; however, precision of force sensation and steadiness of force exertion remains unaltered. This might indicate that control of muscle force is preserved.

Cerebral oxygen desaturation during beach chair position.

CONTEXT: Cases of ischaemic brain damage have been reported in relatively healthy patients undergoing shoulder surgery in the beach chair position. Unrecognised cerebral hypoperfusion may have contributed to these catastrophic events, indicating that routine anaesthesia monitoring may not suffice. Near-infrared spectroscopy (NIRS) provides a non-invasive, continuous method to measure regional cerebral oxygen saturation (rScO2). OBJECTIVES: The aim of this clinical investigation was to evaluate the prevalence of regional cerebral oxygen desaturation in patients undergoing shoulder surgery in the upright position during routine anaesthesia management. We also aimed to identify some causal factors for cerebral desaturation. DESIGN: Prospective, observational, blinded study. SETTING: University hospital. Observation period from 19 05 2008 to 26 08 2008. PATIENTS: Twenty consecutive adult patients presenting for elective shoulder surgery under general anaesthesia in the beach chair position were enrolled. Patients with clinically apparent neurological or cognitive dysfunction were excluded. INTERVENTIONS: Routine anaesthesia management and standard monitoring were used. The responsible anaesthesiologist was blinded to the rScO2 data and was not informed about the purpose of the study. MAIN OUTCOME MEASURES: The prevalence of cerebral oxygen desaturation was measured. RESULTS: With beach chair positioning, rScO2 decreased significantly from 79± to 57±9% on the left side and from 77±10 to 59±10% on the right side (P<0.001). A relative decrease in rScO2 of more than 20% occurred in 80% of patients when the beach chair position was adopted. Postural decreases in cerebral oxygenation were related to blood pressure (r=0.60, P=0.007) and end-tidal carbon dioxide concentration (r=0.47, P=0.035). CONCLUSION: The high prevalence of significant cerebral oxygen desaturation during shoulder surgery in the upright position underlines the need for close monitoring. NIRS might constitute a valuable technique to detect cerebral hypoperfusion in this high-risk group of patients.

Three-dimensional analysis of the orientation and the inclination of the rotator cuff footprint.

HYPOTHESIS: The tuberosities of the proximal humerus relate geometrically to the humeral head in an anteroposterior symmetry. MATERIAL AND METHODS: Twenty-eight cadaveric shoulders were scanned with computed tomography and reconstructed digitally in 3 dimensions. On both tuberosities, 5 facets were identified. An orthogonal planar system using the center of the humeral head as its origin was created to calculate the coordinates of the centre of each facet. The angular position of the center of each facet was measured with reference to the sagittal plane. The inclination of each facet was measured to the axial plane. RESULTS: The presence of 5 distinct facets with a different inclination was confirmed (lesser tuberosity: inferior facet: 77.8° [sd 7.8°]; superior facet: 50.3° [sd 9.3°] - greater tuberosity: superior facet: 20.4° (sd 5.6°); middle facet: 48.6° (sd 5.2°); inferior facet: 92.7° [sd 7.7°]). The angular position of the centers of the facets was less variable (lesser tuberosity: inferior facet: 22.6° [sd 4.3°]; superior facet: 39.2° [sd 4.4°] - greater tuberosity: superior facet: 89.9° [sd 5.9°]; middle facet: 131.6° [sd 6.1°]; inferior facet: 155.1° [sd 6.8°]). CONCLUSION: Our study contributes to the knowledge of the anatomy of the proximal humerus, more specifically of the position of the rotator cuff insertion in relation to the humeral head. These insights are useful in the development of shoulder trauma prostheses.

Prosthetic overhang is the most effective way to prevent scapular conflict in a reverse total shoulder prosthesis.

BACKGROUND AND PURPOSE: Despite good clinical results with the reverse total shoulder arthroplasty, inferior scapular notching remains a concern. We evaluated 6 different solutions to overcome the problem of scapular notching. METHODS: An average and a "worst case scenario" shape in A-P view in a 2-D computer model of a scapula was created, using data from 200 "normal" scapulae, so that the position of the glenoid and humeral component could be changed as well as design features such as depth of the polyethylene insert, the size of glenosphere, the position of the center of rotation, and downward glenoid inclination. The model calculated the maximum adduction (notch angle) in the scapular plane when the cup of the humeral component was in conflict with the scapula. RESULTS: A change in humeral neck shaft inclination from 155° to 145° gave a 10° gain in notch angle. A change in cup depth from 8 mm to 5 mm gave a gain of 12°. With no inferior prosthetic overhang, a lateralization of the center of rotation from 0 mm to 5 mm gained 16°. With an inferior overhang of only 1 mm, no effect of lateralizing the center of rotation was noted. Downward glenoid inclination of 0º to 10º gained 10°. A change in glenosphere radius from 18 mm to 21 mm gained 31° due to the inferior overhang created by the increase in glenosphere. A prosthetic overhang to the bone from 0 mm to 5 mm gained 39°. INTERPRETATION: Of all 6 solutions tested, the prosthetic overhang created the biggest gain in notch angle and this should be considered when designing the reverse arthroplasty and defining optimal surgical technique.

Reliability of the glenoid plane.

HYPOTHESIS: The purpose of this study was to investigate the 3-dimensional (3-D) orientation of the glenoid and scapular planes. Different definitions of the glenoid plane were used and different planes measured, and we hypothesed that the 3-D plane with the least variation would be best to define the most reliable glenoid plane. METHODS: We studied 150 CT scans from nonpathological shoulders from patients between 18 and 80. The scapular plane and 5 different glenoid planes were determined: inferior, anterior, posterior, superior, and neutral. All plane versions and inclination angles were measured. Because all examinations were done in a standardized position to the coronal, sagittal, and transverse planes of the body, the scapular plane could be defined versus the coronal, sagittal, and transverse planes of the body. RESULTS: The version (mean, 3.76) of the inferior glenoid plane showed a significantly lower standard deviation than the version of the anterior (P < .001), posterior (P=.001), and superior (P=.001) glenoid plane (ANOVA). For inclination all planes have a similar variance. The scapular plane was different between gender (P=.022) and correlated with age. CONCLUSION: This study showed that the retroversion of the inferior glenoid is reasonably constant. The osseous anthropometry of the inferior glenoid can offer a reproducible point of reference to be used in prosthetic surgery of the shoulder.

Prosthetic component relationship of the reverse Delta III total shoulder prosthesis in the transverse plane of the body.

The Delta III (DePuy International Ltd, Leeds, UK) reverse total shoulder prosthesis has provided a successful functional outcome in cuff tear arthropathy (CTA); however, internal and external rotation remain compromised. Positioning of the prosthetic components in the transverse plane has theoretically been suggested to affect rotation. Twenty-seven patients who received a Delta III reversed total shoulder prosthesis for CTA were analyzed (mean follow-up, 43 months) using standard radiographs and computed tomography. The position of the prosthetic components and the possible influence of scapular rotation was analyzed using a uniform spatial reference system using axes in reference to the sagittal or coronal plane. We assessed impingement of the humeral component on the glenoid neck in neutral and internal rotation. An increase in the anterior divergence of the glenoid and humeral prosthetic components correlates with an increase in radiologically measured internal rotation (r = 0.932, P < .001). The uniformity of the reference system used seems to allow accurate positioning of the components intraoperatively and can be useful for analysis of the prosthetic component relationship postoperatively.

Consequences of scapular anatomy for reversed total shoulder arthroplasty.

The reverse total shoulder prosthesis provides successful functional outcome in many patients with rotator cuff tear arthropathy. However, scapular notching, a direct consequence of mechanical impingement between the humeral prosthesis and the glenoid, remains a major concern. We presumed a better knowledge of the anatomy of the scapula would enable design or placement modifications to minimize this phenomenon. After establishing a uniform spatial reference system using easy locatable surgical reference points and planes, we analyzed 200 dry bony scapulae and defined the glenoid and infraglenoid anatomy relative to the reference system. The bony rim of the two inferior quadrants of the glenoid forms a semicircle the center of which can be used perioperatively as an easy locatable bony reference point. The infraglenoid tubercle varies in width and length, and can interfere with the humeral part of the reverse prosthesis, creating scapular notching. To avoid notching, we suggest using a convex base plate with a smaller radius than currently used, placing it as low as possible with a 42-mm glenosphere eccentrically assembled to create a posterior offset. If prosthetic overhang cannot be obtained, we suggest removing part of the infraglenoid tubercle.